Computer vision could help Walmart offer a cashierless shopping experience (WMT, AMZN)

This story was delivered to Business Insider Intelligence “E-Commerce Briefing” subscribers hours before appearing on Business Insider. To be the first to know, please click here.

Walmart Executive Vice President and CTO Jeremy King said he was excited about the potential of computer vision technology in stores at the National Retail Federation’s (NRF) annual Big Show, according to Business Insider.

Business Insider Intelligence

The technology can track consumers and products within a store, enabling shopping experiences without any physical checkout process, similar to an Amazon Go store. King didn’t say that Walmart is currently testing computer vision technology — it was reportedly developing such tech in late 2017 — but he did note that he foresees computer vision improving over the next couple years.

Computer vision could help Walmart offer a cashierless shopping experience that avoids the problems of its shuttered “Scan & Go” program. With Scan & Go, Walmart customers used a smartphone app to scan their items as they shopped to avoid waiting in checkout lines. The program was canceled this past May, though Walmart subsidiary Sam’s Club still offers the feature.

At the time of its cancellation, a Walmart spokesperson told Business Insider it was because of low participation, and King has added that there were issues ensuring consumers scanned products correctly. This highlights key problems with Scan & Go-like programs: Consumers have to do more labor to avoid lines, and they’re not trained to handle those tasks, negating much of the convenience the programs are supposed to offer. Computer vision, on the other hand, should function so that consumers can do less work when they shop without having to learn new and confusing processes.

This difference shows that, despite consumers’ interest in self-service tools like cashierless shopping, retailers need to determine the right features to offer. Seventy-three percent of consumers were in favor of self-service innovations to better the shopping experience and minimize interactions with employees in 2018, up 10.6% from 2017, according to a survey from SOTI.

Scan & Go’s issues show that just because shoppers want to be in control of their process doesn’t mean they’ll be able to use the feature effectively. With this in mind, retailers should consider what tasks consumers can effectively take on with minimal training and invest heavily in customer education when trying to innovate with self-service.

The ability to scale computer vision solutions is likely the biggest obstacle to Walmart introducing it to stores. The slate of currently operational cashierless stores are all much smaller than Walmart’s locations. This is because the technology is difficult to scale since it needs to be able to track all consumers and products at once, and the bigger a store is, the more there is to track.

Several companies are working to address this issue of scalability — Amazon is reportedly trying to adapt its technology for bigger stores, and startups like Standard Cognition believe they can facilitate scaling for retail partners — so Walmart will need to find its own solution, or work with a company that’s developed its own, to take advantage of autonomous checkout.